With the X38 chipset out in the wild, PCIe 2.0 is finally upon us, offering twice the bandwidth of the original PCIe standard. Correct me if I'm wrong, but I've read in a number of threads on this forum that today's PCIe 1.0 x16 vid cards don't really make full use of the 4GB/s bandwidth that the x16 connection allows for. Thus, any current cards released with a PCIe 2.0 x16 interface (for example, cards from the Radeon HD 3800 series) are going to be wasting about half of their 8GB/s maximum bandwidth. Doesn't this mean that running a PCIe 2.0 x16 card at x8 (on a PCIe 2.0 motherboard) should incur barely any performance loss (if any at all)?
Am I correct then in figuring that if we were to hook up say, four PCIe 2.0 x16 vid cards in CrossFire on a Gigabyte GA-MA790FX-DQ6 (4 PCIe 2.0 x16 slots, 2 @ x16 and 2 @ x8), we'd effectively be running a full-speed PCIe 1.0 x16 QUAD-CrossFire system?
Am I correct then in figuring that if we were to hook up say, four PCIe 2.0 x16 vid cards in CrossFire on a Gigabyte GA-MA790FX-DQ6 (4 PCIe 2.0 x16 slots, 2 @ x16 and 2 @ x8), we'd effectively be running a full-speed PCIe 1.0 x16 QUAD-CrossFire system?